INDEX

Explanations

instances of the word "followed" or related terms indicating sequences or subsequent actions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 21-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L21R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.21.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

hani

-0.16

ify

-0.15

RowAt

-0.15

dig

-0.15

Dig

-0.15

itters

-0.14

èĦ

-0.14

ties

-0.14

Ø§Øµ

-0.14

Ø§ÙĬÙĬ

-0.14

POSITIVE LOGITS

 closely

0.36

 suit

0.33

-up

0.26

 Suit

0.22

 suite

0.21

ingly

0.21

 shortly

0.20

-Up

0.19

suit

0.19

Suit

0.17

Activations Density 0.024%

instances of the word "followed" or related terms indicating sequences or subsequent actions

No Comments

No Known Activations